Sentaurus Device Electromagnetic Wave Solver
4. Troubleshooting Simulations

4.1 Overview
4.2 Unexpected Results
4.3 Simulations Run Nonstop
4.4 Simulator Aborts

Objectives

• To provide troubleshooting help for EMW simulations.

4.1 Overview

During simulations, EMW displays the most important issues in standard output. In Sentaurus Workbench, this is the *.out file. In addition, EMW writes a more detailed report to the log file set in the Globals section with LogFile, by default <commandfilename>.log.

In the next sections, problems that could arise during simulations are discussed.

4.2 Unexpected Results

Did the simulation converge?

If the simulation converged successfully, you should see the following line in the .out file:

Periodic Detector: Termination criterion met after 793 steps.

Is the excitation correct?

You can double-check the excitation angle in the .out file:

Excitation angles in 3d specification: Theta = 90 deg, Phi = 90 deg, Psi =\
   0.0000e+00 deg.
Excitation angles in 2d specification: Theta = 0.0000e+00 deg, Psi = 0.0000e+00\
   deg.

Are the material parameters correct?

In the .log file, you can check all values for the refractive index and extinction coefficient:

 Setting parameters for region "substrate".
---------------------------------------------------
   Complex refractive index: n = 3.939, k = 0.019767

4.3 Simulations Run Nonstop

Did the simulation converge?

If the simulation converged successfully, you should see the following line in the .out file:

Periodic Detector: Termination criterion met after 793 steps.

Is the value of the time step reasonable?

In the .out file, you can find the chosen time step:

Time step set to 3.2810e-17s.

It is typically in the range 1e-18 – 1e17 s. If it is much shorter, you probably have a very short edge. See the Sentaurus Mesh module.

The maximum deviation first decreases but then increases again. What can I do?

You can:

1. From the log file, extract the time steps when the deviation starts to increase.
2. Perform some plotting around this time step to see where the problem is located.
3. Very often, the wave front hits a certain structure or boundary. If it is an absorbing boundary, try to improve the absorption of the particular boundary by increasing Thickness to 200–300 or increasing SigmaMax to 1e6.

4.4 Simulator Aborts

If EMW aborts:

1. Check the .out file for error messages.
2. Remove freshly added features, or try to reduce your command file as much as possible, for example, remove all Plot, Extractor, and Sensor sections.